README.md

metast-tree

LFP analysis and classification of metastases

Load LFP data

The MATLAB function [trainingData, testData, trainingLines, testLines] = process_data(trainingDataRaw, testDataRaw).

• Inputs

  • trainingDataRaw: table loaded from training_data.xlsx
  • testDataRaw: table loaded from test_data.xlsx

• Output

  • trainingLines: name of the animal line of each row of the training sessions
  • testLines: name of the animal line of each row of the test sessions

Decision trees

Already trained decision trees are loaded in decision-trees/decision_trees.mat. They can be used to predict new data into sham, breast, melanoma or lung categories can be achieved with:

prediction = decisionTrees{iModel}.predict(trainingData.X);

Confusion matrix

The MATLAB function confusion_matrix(ytrue, ypred, classNames, <optional>) computes and plots a confusion matrix with predictions of all models.

• Mandatory Inputs

  • ytrue: N x 1 vector of true classes
  • ypred: N x #models vector of predicted classes. Each column is prediction from a particular model
  • classNames: names of classes (e.g. {sham, breast, melanoma, lung})

• Optional Inputs

  • title: plot title. None by default
  • cLims: color limits. Non by default
  • plotText: boolean indicating whether to show confusion matrix numbers. True by default
  • saveName: complete path for saving the confusion matrix. images/confusion_matrix.png by default

• Optional Outputs

  • confMat: #classes x #classes confusion matrix.

Output example